Herpes Infections and Dementia: Rebutting Alternative Fact
Total Page：16
File Type：pdf, Size：1020Kb
Neurotherapeutics (2019) 16:176–179 https://doi.org/10.1007/s13311-018-00700-5 CURRENT PERSPECTIVES Herpes Infections and Dementia: Rebutting Alternative Fact Richard Lathe1 & Nian-Sheng Tzeng2 & Ruth Itzhaki3,4 Published online: 7 January 2019 # The Author(s) 2018 Abstract Recent commentary in Neurotherapeutics by Nath critically addresses the earlier report by Tzeng et al. that aggressive antiviral treatment (AVT) against herpes simplex virus (HSV) was associated with a later decrease in the incidence of Alzheimer’sdisease (AD). Nath raises issues that we respond to: we point out that (i) the treated group (probably with severe infection) is likely to harbor genetic risk alleles that predispose to both AD and HSV infection—the potential treatment bias cited by Nath would support (rather than challenge) the preventive effect of AVT; (ii) HSV is well known to establish persistent infection in the brain; and (iii) current AVTcompounds used to combat herpes viruses are highly specific for this class of viruses. Instead of “alternative fact,” the findings of Tzeng et al. argue in favor of clinical trials of AVT in AD. Keywords Alzheimer's disease . Herpes simplex virus . Acyclovir . APOE The recent concise commentary by Avindra Nath [1] discusses commentary, Dr Nath queries the results of this study (“alter- the report by one of us (Tzeng et al. [2]), in a recent issue of native fact”), and concludes that it might merely be safe to Neurotherapeutics that addressed a potential protective role treat “patients who develop herpes virus infections with ap- for antiviral therapy (AVT) in Alzheimer’s disease (AD). In propriate antiviral drugs” [1], a proposal wisely consistent this retrospective study employing the Taiwan National with current practice. Although the commentary makes inter- Health Insurance Database, Tzeng et al. reported that short- esting points, other issues appear to be misconstrued and war- term AVT of patients with overt herpes virus infections was rant urgent rectification, which we offer in the hope of clari- followed by a long-term reduction in the subsequent incidence fying the situation. of AD over the following decade ([2]; discussed in [3]). In his Dr Nath is correct is in pointing out that the Tzeng et al. study is retrospective in nature, and thus raises issues concerning prop- Electronic supplementary material The online version of this article er diagnosis in the absence of neuroimaging or postmortem path- (https://doi.org/10.1007/s13311-018-00700-5) contains supplementary ological analysis. However, the Taiwan database employs an material, which is available to authorized users. accredited diagnostic instrument (International Classification of Disease, Ninth Revision, Clinical Modification; ICD-9-CM), and * Richard Lathe thus is unlikely to be a major confounding factor. [email protected] Dr Nath further points out that there was also a lack of * Nian-Sheng Tzeng identification of genetic risk factors. We agree that this latter [email protected] is extremely important because diverse immune-related genes * Ruth Itzhaki are associated both with infection and AD development. [email protected] Centrally, ε4 alleles of the APOE gene are susceptibility fac- 1 Division of Infection and Pathway Medicine, University of tors not only for AD but also for several infectious diseases. Edinburgh, Little France, Edinburgh, UK The ε4 allele is known to increase susceptibility to infection 2 Department of Psychiatry, Tri-Service General Hospital, School of damage by, among others, herpes simplex virus type 1 Medicine, National Defense Medical Center, Taipei, Taiwan, (HSV1) [4, 5], HSV2 [6, 7], Chlamydia pneumoniae [8], Republic of China Chlamydia-associated arthritis [9], and HIV [10]. Some other 3 Nuffield Department of Clinical Neurosciences, John Radcliffe AD risk alleles are also known to predispose to infectious Hospital, Oxford, UK disease; for example, genetic variants in PILRA, a receptor 4 School of Biological Sciences, University of Manchester, Oxford for HSV1 glycoprotein B (that governs HSV1 viral entry into Road, Manchester, UK cells), are associated with AD [11]. Therefore, the population Herpes Infections and Dementia: Rebutting Alternative Fact 177 studied in the Tzeng et al. report (acute HSV1/2 infection) was abundantly expressed in the hippocampus [ 28],asiteofearly likely to have been enriched in AD risk alleles, including degeneration in AD. Indeed, HSV1 DNA can be detected in APOE ε4, and thus may have been inherently at increased risk several brain regions including the hippocampus, both of control of later AD development. No genomic studies were performed individuals and AD patients [29, 30], but levels are upregulated on patients in the Tzeng et al. study, and although we do not in AD brain, as recently confirmed for HSV1/2 as well as for believe that this detracts from the potential protective role of human herpes viruses (HHV) 6 and 7 [31]. The pathological AVT reported by Tzeng et al., we agree (as discussed further signature of AD, Abeta protein, is induced by HSV1 infection below) that such genetic predisposition could potentially un- [32] and is increasingly recognized as an antimicrobial factor derlie positive associations (possibly circumstantial) between ([33], reviewed in [34]) that exerts potent antiHSV1 effects [35, AD and different infectious conditions including herpes zoster 36]. Although de novo brain invasion with HHV following dis- ophthalmicus [12], varicella zoster [13], periodontal disease ease onset cannot be ruled out [37], this is not the case for HSV1 [14–16], Borrelia burgdorferi [17], and Chlamydophila that is widely detected in both AD and control brain tissues [30, pneumoniae [18]. 31, 38]. There are also suggestions that a declining immune We highlight this genetic factor because it must be carefully system in the elderly might predispose to virus reactivation: the borne in mind when considering the next caveat listed by Nath immune system undergoes systemic changes over a lifetime [1]—“selection bias for treatment.” The Tzeng et al. [2]study (e.g., [39]) and reactivation of HSV1 has been reported to take compared patients with acute HSV-induced lesions (predom- place as a function of age [40, 41]. Diverse types of evidence for inantly orolabial and genital) who received AVT versus those reactivation of HSV1 in human brain have been described in who were not treated. It could be argued that the patients who detail [42]. The available data thus manifestly rebut Nath’sview received AVT were likely to have suffered from more severe that “a chronic or persistent infection in the brain ... would be lesions than those who went untreated. In consequence, the highly unlikely.” AVT group (likely with more serious lesions) may have been This indirectly answers the question raised by Nath—“why biased towards inclusion of individuals with a genetic predis- would herpes virus infection in the genitalia or lips lead to de- position to AD. mentia”—in short, it may not (or not directly). Instead, reactiva- Our central argument, therefore, is that the AVT group might tion of epithelial HSV1 might be accompanied (particularly in have been at increased (rather than decreased) risk of later AD genetically predisposed individuals) by reactivation in other tis- development. By contrast, it was precisely the AVT group that sues including the brain, that then initiates the lesions that, over showed a dramatic reduction in later AD development. The the following decades, culminate in clinical AD. Alternatively potential treatment bias cited by Nath argues in exactly the (or in addition), in cases of de novo orogenital infections, in- reverse direction. Thus, it is even more remarkable that their creased peripheral virus load and/or inflammation might increase treatment with AVT apparently prevented most (ca 90%) cases the likelihood that virus enters the brain [3]. of AD development in the decade that followed. Treatment A final issue raised by Nath concerns the apparent bias, if it exists, would therefore entirely support (rather than nonspecificity of the association between infection, AVT, and challenge) the reality of the preventive effect of AVT. AD. He very properly highlights that several types of infectious An additional major issue in Nath’s commentary also calls agent have been associated with AD, and that minor infections of for clarification. He states that “one would have to postulate the elderly can be associated with cognitive decline. However, that the virus establishes a chronic or persistent infection in the this nonspecificity does not extend to AVT. Current AVT com- brain to cause cognitive deficits. However, given what we pounds used to combat herpes viruses are highly specific for this know about these viruses, this would be highly unlikely” class of viruses: these agents (e.g., acyclovir, ganciclovir, [1]. We must point out that this is not correct. penciclovir) are nucleoside analogs that require phosphorylation Seropositivity for HSV1 and HSV2 increases over our life- by the viral thymidine kinase enzyme (and recognition of the time: HSV infection is generally acquired during postnatal life, ensuing triphosphates by viral DNA polymerases) for inhibition and HSV seropositivity increases with age in the USA [19]and of virus proliferation. These compounds are poorly recognized Europe [20], as well as in Taiwan [21], and by age 50 years, the by the corresponding enzymes of other virus classes and display majority of the population harbors HSV1 and/or HSV2 (often in much less or no antiviral activity against them (e.g., vaccinia addition to other herpes viruses such as cytomegalovirus and virus or even veterinary herpes viruses) [43, 44]. There is evi- Epstein–Barr virus). Following primary infection (generally of dence that ganciclovir does have inhibitory activity against ade- epithelial cells) that can be symptomatic or asymptomatic, HSV1 novirus [45], but this is not the first-line antiherpetic drug, and enters a silent (latent) state (reviewed in [22, 23]) principally in others in the class are ineffective against adenoviruses.